WCAG-Aligned Language Tagging for EPUB
WCAG-Aligned Language Tagging for EPUB
Applying WCAG-aligned language tagging across an EPUB catalog so screen readers pronounce foreign-language terms correctly, without over-tagging familiar loanwords.
The Challenge
A publisher needed every non-English word in an EPUB file tagged with the correct language code to meet WCAG accessibility guidelines, so screen readers and other assistive technologies would pronounce and interpret each term correctly. Getting this right took more than technical execution. It took contextual linguistic judgment.
The Partnership
Six Red Marbles worked closely with the publisher to understand both the accessibility requirements and the reading experience they were protecting, keeping technical execution aligned with WCAG standards and editorial intent throughout.
The Strategy
Six Red Marbles built an AI-assisted model to identify non-English words and apply the right language attributes, refining the approach as new edge cases surfaced.
- ✓Detect non-English terms automatically using an AI-assisted identification model.
- ✓Apply WCAG-compliant language tags to true foreign-language content.
- ✓Exclude common English loanwords from tagging, so reading flow stayed natural.
- ✓Layer in structured human review to validate every tagging decision.
The result balanced automation with the editorial judgment WCAG compliance actually depends on.
What Six Red Marbles Designed and Delivered
To avoid over-tagging, Six Red Marbles refined the detection logic mid-project and rebuilt the files around it:
- AI-assisted detection model tuned to the publisher's content
- Differentiated logic separating true foreign-language terms from accepted loanwords
- WCAG-compliant language tagging applied across the EPUB file
- Updated files aligned with the publisher's editorial expectations
The Learning Experience
Screen reader and assistive technology users now hear foreign-language words pronounced and interpreted correctly throughout the text, while familiar loanwords read the way they always have. The accessibility work stays invisible to the reading experience itself.
The Impact
- ✓WCAG-aligned language tagging implemented in full across the file
- ✓Over-tagging issues identified and resolved before delivery
- ✓Final files fully aligned with both accessibility standards and editorial expectations
[PENDING: quantifiable metrics for this engagement — e.g., number of terms or instances tagged, languages covered — were not available in source material. Add a stats row here if the submitter can provide them.]
Why It Matters
As accessibility standards evolve, publishers need workflows that pair automation with editorial judgment. This case shows how refining a tagging model in place, rather than applying it rigidly from the start, keeps compliance and readability intact together.
